Fenstermarcher Gary_The Knower and the Known

8/13/2019 Fenstermarcher Gary_The Knower and the Known

1/61

THE KNOWER AND THE KNOWN: THE NATURE OF

KNOWLEDGE IN RESEARCH ON TEACHING*

Gary D Fenstermacher

University of Arizona, Tucson

I. INTRODUCTIONThis chapter is a review of conceptions of knowledge as they appear in

selected bodies of research on teaching. Writing as a philosopher of education,my interest is in how notions of knowledge are used and analyzed in a number ofresearch programs that study teachers and their teaching. Of particular interest isthe growing research literature on the knowledge that teachers generate as aresult of their experience as teachers, in contrast to the knowledge of teaching

that is generated by those who specialize in research on teaching. Thisdistinction, as will become apparent, is one that divides more conventionalscientific approaches to the study of teaching from what might be thought of asalternative approaches.

A number of good reviews of the teacher knowledge literature are availableelsewhere. Though these tend to be confined to a particular genre of teacherknowledge research, they are thoughtful, probing and helpful. Among them areKathy Carter's (1990) chapter in the Handbook of Research on TeacherEducation, Alan Tom and Linda Valli's (1990) philosophically grounded reviewof professional knowledge; and Peter Grimmett and Allan MacKinnon's (1992)

extensive analysis of craft conceptions of teaching in a previous volume in thisseries. What distinguishes the present review from these others is that it seeks tobe fairly inclusive of the teacher knowledge literature, though restrictive in itsanalytical categories. I shall examine a number of different research programsthat either explicitly purport to be about teacher knowledge or that expand whatis known about teaching. The examination, however, will be restricted to the____________*A significant portion of this chapter is based on prior work done in collaboration with Virginia

Richardson. This work resulted in an earlier, substantially different version of this chapter,

presented at a conference on teacher knowledge in Tel Aviv, Israel, in June, 1993. Frederick

Ellett, Peter Grimmett, Hugh Munby, Nel Noddings, Robert Orton, and Denis Phillips, offered

extended and extremely helpful critiques of this chapter, for which I am deeply indebted. For

the many errors of fact and interpretation from which they have spared me, I thank them. They

bear no responsibility for what remains. Peter P. Grimmett and Lee S. Shulman were theeditorial consultants for this chapter, a role they performed with diligence and grace.

This text is the manuscript version of the following published chapter: Gary D Fenstermacher(1994), The Knower and the Known: The Nature of Knowledge in Research on Teaching, in LindaDarling Hammond (Ed.), Review of Research in Education, 20(pp. 3-56), Washington,DC:: American Educational Research Association. The text of this manuscript version may not beidentical to the published piece; the pagination will definitely be different..


2/61

2

epistemological aspects of these research programs. By epistemological aspects,I mean those features of the research that assert or imply notions about thenature of knowledge: What forms it takes, how it is justified, how it is

differentiated from such related concepts as belief and opinion, and how it figuresinto different conceptions of science and human reasoning.

One may with reason wonder whether any benefit is to be gained byworking one's way through the epistemological underbrush while traversing theteacher knowledge terrain. The value is in having a better understanding of whatis involved in researchers' claims to know something about teaching as well astheir claims that teachers know some things about teaching. These claims canand do become the basis for educational policy, even when our acceptance ofthem is more a matter of ideological tilt or cosmetic appeal than of clear-headedanalysis and reasoned deliberation. Take, for example, the frequently lauded

"knowledge base" for teaching.

In the United States, many members of the policy making community areembracing a view of teacher knowledge and skill that represents a limitedepistemological perspective on what teachers should know and be able to do.This perspective is classified in this review as the Formal perspective. It isgrounded in a conception of the social and behavioral sciences that arethemselves constructed isomorphically to the physical sciences. It is from thisperspective that we have built the much-vaunted "knowledge base" for teaching.It is this knowledge base that in turn gives rise to such policy initiatives asnational certification for teachers, accountability and performance assessment in

teaching, research-based programs of teacher education as well as research-based designs for the accreditation of teacher education, and some (though notall) of the current initiatives in the development of subject field, grade level, andstate level standards for student learning.

If our educational policies are defended on the basis of their foundations inscience, and science in turn rests on epistemology for an understanding of thenature and scope of its knowledge claims, then epistemology is crucial. Supposeit turned out that the epistemology was faulty, or that it was more limited andconstrained than previously thought, or that it was but one among a number ofpossible epistemologies, or that it was the wrong epistemology for the nature of

the inquiry. What then? If educational policy is grounded in weak or erroneousassumptions about the nature of knowledge, there is a high likelihood that it willfail to address the problems and aspirations of education in positive andameliorative ways.

This review is an occasion to reconsider the epistemological character ofwhat is and can be known by and about teachers and about teaching. It begins,in part II, with a description of a number of different research programs in the


3/61

3

study of teaching. These programs are grouped in ways that open useful avenuesinto issues of epistemology. After examining the research literature, we turn, inpart III, to an analysis of its epistemological structure. Of particular interest here

is the kind of knowledge that researchers and teachers claim to have, and how weknow they have it. This inquiry into the nature of teacher knowledge raises issuesbeyond the confines of epistemology, leading us to look, in part IV, at conceptionsof science that undergird the various forms of teacher knowledge research. Theinquiry into science shows both promise and problems for gaining firmerepistemological ground. The problems lead to a search for other ways ofestablishing the knowledge that researchers claim they or the teachers they studyhave. Part V introduces the topic of practical reasoning in teaching, with an eyetoward recasting portions of the teacher knowledge research as research onteacher reasoning. The chapter concludes with an assessment of what needs tobe done to continue the high potential of newer research programs for the study

of teacher knowledge.

Throughout this review consideration is given to the knower and theknown in teacher knowledge research. Questions will be raised about who is theknower, and what it is that the knower knows. Is the knower the researcher(typically thought of as a university professor), or the teacher, or both researcherand teacher? Is whatever it is that is known known only by the researcher, onlythe teacher, or by both? If both teacher and researcher are considered seekersand producers of knowledge, how are they different with respect to the researchendeavor? These are the questions that come to mind as one pursues the severalliteratures on the subject of teacher knowledge. We turn now to these literatures,

examining them in ways that address these questions about the knower and theknown.

II. AN OVERVIEW OF THE TEACHER KNOWLEDGE LITERATUREAfter struggling with a number of devices for arraying the literature in

ways that would facilitate epistemological scrutiny, a series of questions offeredthe best approach to setting forth the relevant research. There are four questions,as follows:

A. What is known about effective teaching?B. What do teachers know?C. What knowledge is essential for teaching?D. Who produces knowledge about teaching?

These four questions do not exhaust the research studies dealing with knowledgein teaching. Instead they prompt consideration of those research initiatives thatraise substantial epistemological questions. Hence many studies that, on their


4/61

4

face, appear relevant to the topic are missing from this review. That is becausethe point of the review is to grapple with research that raises substantial, and onoccasion, controversial issues in epistemology. Thus the work included here is

foundational to a particular research program, it is exemplary of that program, orit is a very recent and extensive study within a particular approach to research onteaching.

Some foreshadowing of the answers to these four questions will provehelpful in following the overview. The question, What is known abouteffective teaching?permits us to address the concept of knowledge as itappears in standard or conventional behavioral science research. In this chapter,this concept of knowledge is called Formal knowledge, and it is abbreviated TK/F(Teacher Knowledge: Formal). The process-product studies of teaching areperhaps the most well-known instance of this form of knowledge. The second

question, What do teachers know?points to research that seeks tounderstand what teachers know as a result of their experience as teachers. Quitea few kinds of knowledge are suggested in answer to this question, includingpractical, personal practical, situated, local, relational, and tacit. These types ofknowledge are designated TK/P, for "Teacher Knowledge: Practical."

These first two questions encompass the epistemological types ofknowledge suggested in the research literature, but they do not exhaust the rangeof the literature itself. The third question, What knowledge is essentialfor teaching?directs us to the research program of Lee Shulman and hiscolleagues. It will be argued that Shulman's work does not introduce types of

knowledge different from those represented in answers to the first two questions,but rather seeks to show what forms and types of knowledge are required inorder to teach competently. The fourth question, Who producesknowledge about teaching?permits us to address the difference betweenknowledge generated by university-based researchers and that generated bypracticing teachers. The work of Marilyn Cochran-Smith and Susan Lytle is quiteprominent in this category, and thus serves as the exemplar in addressing thefourth question.

Before turning to the specific questions, it is necessary to attend briefly tomatters of definition. In this review I frequently refer to research programs. This

expression typically refers to Imre Lakatos's (1970) notion of a broad group ofresearch studies united by common methodological rules. In this review, theexpression refers to the work of certain researchers, and the studies based on thework of these researchers. Thus, for example, I will refer to Michael Connelly andJean Clandinin's work on personal practical knowledge as a research program,and in doing so I mean to include within the category their own work on thisconcept as well as the work of collaborators and like-minded colleagues. Henceall research studies that use Connelly and Clandinin's notions of personal practical


5/61

5

knowledge as their core concept are a part of this research program. A similargrouping is made for Schn's work on reflective practice, Shulman's work on typesof knowledge about teaching, and Cochran-Smith and Lytle's work on the teacher

researcher.

The second definitional matter is the notion of knowledge type. A surveyof the teacher knowledge literature turns up a host of knowledge names.Examples include strategic knowledge, propositional knowledge, relationalknowledge, craft knowledge, local knowledge, case knowledge, situatedknowledge, tacit knowledge, personal knowledge, and so forth. All these namesdo not necessarily pick out different typesof knowledge. These knowledgenames are, in some ways, like the names we give to people. Each of us has agiven (formal) first name, usually a nickname, and often another name that is afamily favorite; we may have still other names because of our membership in

certain religious or social groups. These names all pick out the same person,though their context of use is different. While this aspect of naming is easilyunderstood in connection with persons, it can be hard to follow with ideas orconcepts. For this reason I restrict the term typeto discrete epistemologicalcategories, allowing the knowledge names to flower according to the preferencesand inclinations of the researchers who coin, adopt, or adapt them. The notionsof formal and practical knowledge mentioned above are instances of typesofknowledge. More will be made of this point in the section that follows theoverview; enough has been made of it here to carry us through the currentsection.

A. What is known about effective teaching?This first question permits us to consider a key approach to the topic of

teacher knowledge, that of the social and behavioral sciences (hereafter theexpression "social sciences" includes the social and behavioral sciences). Thoughthe question reads, What is known about effective teaching?it couldas readily have been framed as, What is known about successfulteachers?or What is known about what makes teachers good atwhat they do? The answer to these questions embraces all of the researchthat deals with relationships between or among variables, including nearly all ofthe process-product research, as well as a portion of the research pertaining toteacher thinking, cognitive processing, teacher expectancy, as well as a number

of studies dealing with the topics of learning to teach and staff development.

This body of research does not explicitly mention teacher knowledge; itdoes not specifically identify itself as teacher knowledge research. Yet researchprograms that study relationships between variables entail a conception ofknowledge about teachers and teaching that some believe to be critical for theadvancement of the field. Researchers in this category do not see themselves as


6/61

6

studying teacher knowledge so much as they perceive themselves producingknowledge about teaching. Using methods and designs found in the socialsciences, they seek the determinants of good (successful, effective) teaching.

Their work rests on a belief that if their methods and designs are in accord withaccepted scientific theory and practice, their results may safely be accepted asknowledge about teachers and teaching.

This point of view is perhaps nowhere stated with more clarity andsuccinctness than in the work of N. L. Gage. In The Scientific Basis of theArt of Teaching, Gage (1978) sets forth his view of the relationship betweenscience and teaching. The title of the book is carefully chosen, for Gage does notcontend that teaching can or should be a science, but is rather an art based on ascience. The science that Gage believes is the basis for teaching is psychology,particularly as practiced by university-trained research psychologists. In an

update to The Scientific Basis, this one entitled Hard Gains in the SoftSciences: The Case of Pedagogy, Gage (1985, p. 7) states that "ascientific basis consists of scientifically developed knowledge about therelationship between variables." Working within what might be called a standardor conventional conception of science, Gage argues that scientific knowledge isnomothetic (law-like), "it holds for the more or less general case under specifiedconditions with all, or all but a few, other variables held constant" (1985, p. 8).On how research gains its warrant, Gage (1985, p. 8) states that "nomotheticknowledge becomes more trustworthy as it is confirmed through replications, thatis, repetitions of more or less similar investigations. Such confirmations add moreto the persuasiveness of the evidence than does the statistical significance, no

matter how strong, of a single result."

In a volume honoring Gage's contributions to research on teaching, DavidBerliner contended that "we are on the threshold of creating a scientific basis forthe art of teaching that will be acceptable to the general public as trulyspecialized knowledge" (Berliner, 1987, p. 6, emphasis in original). Berlinergoes on to argue that educational science has made practical contributions toeducation, among them "success rate, structuring, academic feedback,monitoring, motivating, expectancy, and wait time" (p. 18). In another effort tocodify the knowledge accumulated through scientific studies for use by personspreparing to teach, the American Association of Colleges for Teacher Education

commissioned a volume entitled, Knowledge Base for the BeginningTeacher(Reynolds, 1989). In his preface to the book, William Gardner wrote,"this book seeks to demonstrate that teaching does have a distinctive knowledgebase .... This knowledge base has been generated in research ..." (Gardner, 1989,p. ix).

The AACTE Knowledge Base Project was an effort to draw together whatwas known at the time about effective or successful teaching, for use by teachers


7/61

7

and teacher educators. Two other projects were also completed in the late1980's, both of them with a distinctive emphasis on the scientific understandingof teaching. The first of these was The International Encyclopedia of

Teaching and Teacher Education(Dunkin, 1987), where many of thebest known researchers on teaching summarized their work to date. In explainingthe nature of this work, Michael Dunkin (1987, p. xiv) states:

"Substantive knowledge" as used here refers to information aboutteaching and teacher education obtained as a result of attempts totest hypotheses and answer questions about these two processes,their relationships with each other, and with other phenomena.Knowledge about teaching methods and other classroomoccurrences is at the core of this Encyclopedia.

The other initiative was the third edition of the Handbook of Research onTeaching(Wittrock, 1986), a volume that many regard as the definitivecompilation of what the American educational research community knows aboutteaching.

The science-oriented approach characteristic of these publications rests ona particular conception of knowledge. This conception is a modification of what isknown as the standard, or justified true believe account of human knowledge. Itis referred to here as Formal knowledge, and abbreviated TK/F. This knowledge isgained from studies of teaching that employ conventional scientific methods,quantitative and qualitative, where said methods and accompanying designs are

intended to yield a commonly accepted degree of significance, validity,generalizability, and intersubjectivity. This formal conception of knowledge isdifferent from the type of knowledge that characterizes research programsdesigned to answer the next question.

B. What do teachers know?This question arises in the case of those researchers who want to know

what teachers already know, in contrast to producing knowledge for teachers touse. A critical presupposition of research in this category is that teachers know agreat deal as a result of their training and experience. A major contributor to thiscategory of research, Clandinin (1986, pp. 8-9), contends that "teachers are

commonly acknowledged as having had experiences but they are credited withlittle knowledge gained from that experience. The omission is due in part to thefact that we have not had ways of thinking about this practical knowledge and inpart because we fail to recognize more practically oriented knowledge."

For purposes of this review, it is useful to divide into two strands theresearch that seeks to answer the question, What do teachers know? The firststrand encompasses the work of Freema Elbaz and Connelly and Clandinin (and


8/61

8

those research studies that spring from their work). The second strand is basedon the reflective practice notions of Donald Schn, and is most evident in thework of Hugh Munby, Tom Russell, and a number of other Canadian and U. S.

researchers (it is of more than passing interest that a significant portion of theresearch on teachers' practical or reflective knowledge is done in Canada). Bothstrands seek a better understanding of the knowledge teachers bring to their workand the understandings they have of it, although each strand rests on a differenttheoretical and methodological foundation. We will begin with the strandrepresented by Elbaz and Connelly and Clandinin.

Elbaz (1991, 1983) was one of the earliest contributors to this form ofteacher research. Elbaz studied a high school teacher she called "Sarah" over atwo year period, in order to gain a sense of what Elbaz calls her "practicalknowledge." "This knowledge," writes Elbaz (1983, p. 5), "encompasses first hand

experience of student's learning styles, interests, needs, strengths and difficulties,and a repertoire of instructional techniques and classroom management skills."Elbaz contended that the teacher's practical knowledge ranged over five areas,knowledge of self, milieu, subject matter, curriculum development, andinstruction. Moreover such practical knowledge is represented in practice in threeways: First as rules of practice, which are "brief, clearly formulated statements ofwhat to do or how to do it in a particular situation" (1983, p. 132); second, aspractical principles, "more inclusive and less explicit formulations in which theteacher's purposes ... are more clearly evident" (p. 133); and third, as images,which in the case of Sarah are defined as "a brief, descriptive, and sometimesmetaphoric statement which seems to capture some essential aspect of Sarah's

perception of herself, her teaching, her situation in the classroom or her subjectmatter, and which serves to organize her knowledge in the relevant areas" (p.137).

As one reads Elbaz's study of Sarah, it becomes clear that her interest is inwhat Sarah knows or believes about her work, and how Sarah's understanding ofher work might itself be understood. This approach is not that of a conventionalscientific research program, where the focus would be on whether Sarah isinstructionally effective, or on how she thinks given this or that theoreticalorientation, or whether her actions are predicted by some theory. Instead, Elbazseeks to grasp Sarah's knowledge of her working world, without imposing theory

or established methods on the form of inquiry and without structuring Sarah'sresponses within an existing tradition of academic research. In undertaking thiswork, Elbaz finds that what Sarah knows is not theory, or empirical propositions,but how to carry out instructional tasks, resolve conflicts, adjudicate competingconsiderations, and connect aspirations to plans and then to instructionalperformance. Elbaz contends that these understandings make up Sarah's"practical knowledge."


9/61

9

The notion of practical knowledge also figures prominently in the work of F.Michael Connelly and D. Jean Clandinin (Clandinin & Connelly, 1987, 1991;Connelly & Clandinin, 1985. 1988, 1990). They refer to the critical element in

their work as "personal practical knowledge." Clandinin (1992. p. 125) describesthis form of knowledge as follows:

We see personal practical knowledge as in the person's pastexperience, in the person's present mind and body and in theperson's future plans and actions. It is knowledge that reflects theindividual's prior knowledge and acknowledges the contextualnature of that teacher's knowledge. It is a kind of knowledge carvedout of, and shaped by, situations; knowledge that is constructedand reconstructed as we live out our stories and retell and relivethem through processes of reflection.

Those whose predilections are for more formal, more conventionallyscience-oriented, perspectives on the nature of knowledge will experience someconsternation on reading this definition. It can be a challenge to follow Connellyand Clandinin as they work through their complex and elaborate conception ofteacher knowledge. My own understanding of their work is helped by gaining asense of their motivation, which will detain us only a moment. In her first book-length study of teacher knowledge, Clandinin (1986) describes her dissatisfactionas an elementary school counselor, in part because of the way theory andresearch were being imposed on teachers, and because the teachers with whomshe worked had a number of negative experiences with school counselors. "In my

work with teachers, I experienced a personal dissatisfaction with the wayteachers are viewed. The prevailing view and organization of the educationalenterprise give little credit to their knowledge" (Clandinin, 1986, p. 8). Over time,Clandinin came to reject the imposition of academic theory and research onteachers, preferring instead to try to understand how they think about their ownwork and what knowledge they use as a basis for their actions.

Recently Connelly and Clandinin (1990, p. 12) stated that "what is at stakeis less a matter of working theories and ideologies and more a question of theplace of research in the improvement of practice and of how researchers andpractitioners may productively relate to one another." In taking teachers seriously

as holders of practical knowledge, it becomes necessary for Connelly andClandinin to find ways of learning what teachers know without employingmethods that distort, destroy or reconstruct this knowledge. What theydiscovered in undertaking this form of inquiry is similar to Donald Polkinghorne'sinsight when he asked whether clinical practitioners had a better understandingof human psychopathology than researcher-theorists did: "Practitioners work withnarrative knowledge. They are concerned with people's stories: they work withcase histories and use narrative explanations to understand why the people they


10/61

10

work with behave the way they do" (Polkinghorne, 1988, p. x). Connelly andClandinin adopted the techniques of narrative and story (and the related conceptof images) as ways of avoiding the excessive imposition of external theories and

constructs on the personal practical knowledge of teachers. They argue thatinquiries such as theirs must be "soft" or "gentle," else the storytelling urge will bestifled by the absence of trust between teachers and those who seek tounderstand what they know.

Using the notion of images, Clandinin probed the perspectives of theteachers with whom she worked. Clandinin views images as "the coalescence ofa person's personal private and professional experience" (1986, p. 166). An imageis "a way of organizing and reorganizing past experience;" it "embodies a person'sexperience; finds expression in practice; and is the perspective from which newexperience is taken" (1986, p. 166). Images are conceptions that teachers have of

their work that account for what they do. For example, one image is of the schoolas a second home for students, leading the teacher to think that she is a kind ofcustodial parent to her students, in charge of a classroom that is a home-likeenvironment.

The notions of story, image, narrative, narrative unity and embodiedknowledge are all central to the Connelly and Clandinin research program.Though they have taken some pains to explain these concepts (see, for example,Connelly and Clandinin, 1985, 1988, & 1990), they remain puzzling concepts formany persons outside this research program. The notion of embodiedknowledge, for example, is taken from the work of Mark Johnson, a philosopher

who has argued that "human beings have bodies that are the locus of theircomplex interactions with their environment" (M. Johnson, 1989, p. 366; see also,M. Johnson, 1987). Johnson's contention that there are significant connectionsbetween the "structures of our bodily experience and what we regard as our'higher' cognitive capacities" (M. Johnson, 1989, p. 366), while epistemologicallyprovocative, is quite distant from the mainstream of contemporary scholarship inepistemology. The notion of "narrative unity," so strongly emphasized in the workthat Connelly and Clandinin do with teachers, is based on the work of the ethicaltheorist, Alisdair MacIntyre (1984). While Connelly and Clandinin argue that thenotion of narrative unity is central to their program, it remains a difficult conceptto unpack with precision, especially in the context of classroom teaching.

These sentiments concerning the difficulties that can arise when usingnarrative and story as the basis of studying teacher knowledge are echoed inCarter's (1993) and John Willinsky's (1989) constructive critiques of the conceptsand methods employed in this research program, as well as in related programsthat employ techniques of story and narrative. It is possible, I believe, for aresearch program to be so heavily enmeshed in theory and conceptualization thatit risks being blurred by it own abstractions. Connelly and Clandinin face that risk


11/61

11

with their current initiative. On the other hand, this program offers bothconception and method that tap what appears to be an important and frequentlyignored type of human knowledge.

This second type of knowledge is here designated TK/P, or TeacherKnowledge/Practical. Teacher Knowledge/Practical is that knowledge orunderstanding developed from participating in and reflecting on action andexperience. It is bounded by the situation or context in which it arises, and it mayor may not be capable of immediate expression in speech or writing. TK/P isgenerally related to how to do things, or the right place and time to do them, orabout how to see and interpret events related to one's actions. With thisdefinition in place, we turn to the second strand of responses to the question,What do teachers know?

As Donald Schn examines practical knowledge, it is different from theconception put forth by Connelly and Clandinin. Schn grounds his work in thepragmatism of Dewey, and in his own studies of architects, psychotherapists andengineers. This work, set forth most explicitly in The ReflectivePractitioner(Schn, 1983; see also Schn, 1987, 1991), rests on a profounddissatisfaction with what Schn calls Technical Rationality, as well as a search fora better way of understanding how professionals work in action. TechnicalRationality is, in a nutshell, the application of conventional social science to theproblems and tasks of everyday professional practice. It does not work, contendsSchn, because it rests on a fundamental misconception of what professionalsdo. Their knowledge is not the knowledge of science, which occupies "the high,

hard ground," populated with research-based theory, but rather is the knowledgeof practice, found in the "swampy lowland where situations are confusing"messes" incapable of technical solutions" (Schn, 1983, p. 42).

For Schn, our knowing is in our action. Explicating this notion, Schnrefers to such concepts as knowing-in-action, reflecting-in-action, reflecting-in-practice, and framing and reframing. These concepts are the heart of what Schncalls "the epistemology of practice." This notion is intended to set off the conceptof theoretical, formal knowledge from that of practical knowledge. Schn's view isthat not only are the sources of these two kinds of knowledge different, they areepistemologically different types of knowledge. Though I have been critical of the

concept of an epistemology of practice (Fenstermacher, 1988), it now seems tome to be quite consistent with the larger notion of practical knowledge. Thus ifone accepts practical knowledge as a legitimate epistemological type, a matter tobe explored in the next section, the notion of an "epistemology of practice" is notnearly so troubling as it is if one confines oneself to formal knowledge as the sole,legitimate epistemological type.


12/61

12

A number of excellent research studies are grounded in conceptspropounded by Schn. Many of these can be found in The Reflective Turn,edited by Schn (1991); Teachers and Teaching: From Classroom to

Reflection, edited by Tom Russell and Hugh Munby (1992); andEncouraging Reflective Practice in Educationedited by Renee Clift,Robert Houston and Marleen Pugach (1990). Russell and Munby (1992, p. 2)state that "Schn's (1983) distinction between technical rationality and theknowledge of practice has drawn our attention to the significance of theknowledge that teachers acquire from their own experience." In contrast toClandinin and Connelly, Russell and Munby do not argue that the narratives ofteachers are prima facieexamples of teacher knowledge. Moreover, they andother researchers committed to Schn's conception of reflective practice (see, forexample, Erickson and MacKinnon, 1991, and Grimmett and Chehan, 1990) showa greater interest in how knowledge arises in the context of action as well as for

the consequences of this knowledge for practice than in the description of thisknowledge through narrative or story (hence the greater use of video tape torecord the actions of teachers). Many of the researchers in this second strandhave come to their conceptions of teacher knowledge through attempts tounderstand and improve the initial preparation of teachers.

In reading the research studies from both strands, one quickly gains asense of different assumptions, different methods, and different outcomes. Forexample, in the Elbaz/Connelly/Clandinin strand, teacher knowledge appears to beinferred from narratives and stories (perhaps they would prefer to say thatnarrative and story revealthe knowledge that teachers possess), while in the

Schn/Munby/Russell strand, teacher knowledge is inferred from action thatarises in the course of experience as a teacher. Another difference one senses inreading these studies is that Elbaz, Connelly and Clandinin seem prepared toaccept teacher statements, stories and images as knowledge, while Munby andRussell (and like-minded colleagues such as Erickson, Grimmett and MacKinnon)are more cautious about according the status of knowledge to what teachers sayor do.

Researchers in the Schn strand argue, as do Connelly and Clandinin, thatteachers produce and possess their own knowledge, but Schn strandresearchers consider it far more of a task to tease out precisely what knowledge is

involved in action, and how this knowledge is altered in subsequent action.Researchers in the Connelly and Clandinin strand, on the other hand, appear togrant the teacher's insights the status of knowledge simply as a result of theteacher's having given expression to his or her conceptions of what took place inthe classroom (this last comment may be inaccurate, for it is not clear to mewhether the insights and understandings teachers express in narratives areaccorded the status of knowledge merely because they are teacher-


13/61

13

articulated insights and understandings, or whether they must first meet somecategorical standard before being accorded the status of knowledge--thisconfusion may be stated in the form of a question, What is it that the teacher(knower) knows, and how is it known that the teacher knows it?).

Despite the differences between the two strands in approach and method,both strands seek a conception of knowledge arising out of action or experience,that is itself grounded in this same action or experience. Both strands show littlepatience for conventional social science methods, and for the epistemologicalconceptions that inhere to such methods. While their supporting theory andresearch methods differ, it is my sense that they are in pursuit of the same

epistemological type(Jana Noel, 1993, makes a similar point with respect to theconcept of intentionality that spans both these research strands). The Russell andMunby (1992) anthology, for example, includes chapters from both strands ofpractical knowledge. However there is little talk across the strands in the book,either in service to finding common interests or themes, or explication ofimportant differences between the two. Given that both strands seek theexplication of the same epistemological type, and that each uses novel andprovocative methods for eliciting and understanding teacher knowledge, theexplication of the concept of practical knowledge has much to gain from moredialogue between these two strands. Further dialogue is also likely to benefit athird approach to teacher knowledge, one that appears to employ both formal and

practical versions of knowledge.

C. What knowledge is essential for teaching?The research program that answers this question was launched by Lee

Shulman in his wide ranging efforts to understand the nature of knowledgeundergirding the enterprise of teaching (see, for example, Shulman, 1984, 1986a& b, 1987a & b, 1989, 1992). The groundwork for this program is laid in twoimportant articles, one in the Harvard Educational Review(Shulman,1987a) where Shulman set forth an argument for "the content, character, andsources for a knowledge base of teaching" (p. 4). The other is his 1985presidential address to the American Educational Research Association, where he

advanced the notion of pedagogical content knowledge (Shulman, 1986b). In theformer piece, Shulman sets out the broad categories of knowledge required forteaching, while in the latter, he focuses upon a few of the categories dealing withcontent and pedagogical knowledge, contending that these are not nearly so wellunderstood as the other categories of knowledge.

The research program that has emerged from these conceptions of teacherknowledge addresses questions of how students learn to teach subjects that they


14/61

14

already know or are in the process of acquiring (Grossman, 1990; Grossman,Wilson & Shulman, 1989; Wilson, Shulman & Richert, 1987). Much of this workis about what Shulman calls "pedagogical content knowledge." Barbara Nelson

(1992) opened her review of a book by one of Shulman's students with thiscomment: "In his 1985 presidential address, Lee Shulman tossed off the phrase"pedagogical content knowledge" and sparked a small cottage industry devoted tothe scholarly elaboration of the construct ..." (p. 32). While the notion of "tossedoff" seems a bit ungenerous, given the amount of scholarly development thatwent into the concept, there is no doubt that the concept has spawned anextensive set of research studies. What is difficult to ascertain in this work iswhether it is distinctly formal or practical in type, or a blend of the two.

What distinguishes this work from that of, say, Elbaz, is that it is morenormatively oriented. That is, it is rooted in a conception of what teachers should

know and be able to do, with a concern for what categories and types ofknowledge are required to achieve this state of competence (in the larger, morerobust sense of the term). Elbaz, on the other hand, assumes a more descriptivestance, one that she hopes will reveal what teachers know and understand as aresult of training, experience and reflection, not what a teacher shouldknow inorder to be successful or effective. (Note, however, that Elbaz's work has itsnormative dimensions; these are evident in her conception of how teachers oughtto be viewed by those outside the occupation.)

Pamela Grossman (in press) alludes to this difference between the tworesearch programs when she states that "studies of personal practical knowledge

are not intended to inform as much as to illuminate classroom practice fromteachers' perspectives" (p. 14). While Grossman does not so state, the suggestionis that the research program in which she and Shulman are engaged is tilted moretowards informing than illuminating classroom practice. I do not wish to extractmore meaning from this sentence than it was intended to impart, yet it nudgesone further along a thought line that is suggested in reading Shulman's work onteacher knowledge: Formal knowledge has a larger role here than does practicalknowledge, and perhaps even takes precedence over the practical.

It should not be inferred from this comment that Shulman is arguing forthe precedence of experimental or correlational research results when compiling

the knowledge base for teaching. He argues, for example, that, "the results ofresearch on effective teaching, while valuable, are not the sole source of evidenceon which to base a definition of the knowledge base of teaching" (Shulman,1987a, p. 7). However, to say that the primary knowledge type in the Shulmanresearch program is formal is not to restrict the range of this work to boundariesset for experimental or correlational studies. One can range well beyond theseboundaries and still participate in the formalist conception. Shulman appears todo this quite successfully.


15/61

15

There are, however, aspects of the Shulman program that suggest a stronginterest in practical knowledge. In his presidential address, he distinguishes threeforms of knowledge, propositional, case, and strategic knowledge. Case

knowledge "is knowledge of specific, well-documented, and richly describedevents" (Shulman, 1986b, p. 11). Case knowledge is distinguished from casemethods, which are "a particular strategy of pedagogical transformation--astrategy for transforming more propositional forms of knowledge into narrativesthat motivate and educate" (Shulman, 1992, p. 17). This reasoning suggests thatcases and case methods are ways of transforming propositional knowledge intopractical knowledge. Indeed, Shulman (1992, p. 21) argues that "from anepistemological perspective, cases may be more congruent with the forms ofpractical knowledge that undergird the varieties of practice, in teaching and otherfields." There is, as you may have noticed, some confusion here. In the 1985presidential address, Shulman argued that propositional and case knowledge are

different, and that the three types of propositional knowledge are empirical andphilosophical inquiry, practical experience, and moral reasoning. In the 1992article on cases, he argues that case methods are a way to transformpropositional knowledge into case knowledge, the latter being a type of practicalknowledge.

Is this confusion the result of developing different perspectives in the sevenyears between '85 and '92? The span of time surely accounts for some of thechange, for Shulman, like any active scholar, continues to refine conceptsdeveloped in earlier work. Yet development over time is not the sole explanation.Rather, what Shulman has been calling a type of knowledge seems more like a

description of the way a practitioner holds the knowledge he or she has; sheholds it propositionally, or as a story or narrative (case). Unfortunately thisconstruction, while appealing as a way of making sense of propositional and caseknowledge, will not work for strategic knowledge. Strategic knowledge, Shulmancontends, is what we have when principles of practice conflict and we must actto resolve the conflict. This type of knowledge constitutes professional judgment,"the hallmark of any learned profession" (Shulman, 1986b, p. 13). Strategicknowledge appears to be the skilled adjudication of conflicts between the rules orprinciples (developed out of propositional knowledge) and the specific instancesencountered in practice (cases or case knowledge).

Interestingly, strategic knowledge has the closest affinity to the notion ofpractical knowledge as exemplified by the TK/P category. In his description of it,Shulman does not say that strategic knowledge is a way of holding knowledge,while propositional and case knowledge do seem to be a way of holdingknowledge. Both these latter categories, propositional and case, appear to be ablend of formal and practical typesof knowledge, with a heavy dose of theformal, while strategic knowledge appears to fall well within the boundaries ofpractical knowledge.


16/61


17/61

17

The writings of Cochran-Smith and Lytle leave little doubt that teachersshould be among the producers (they often use the term generators) of

knowledge. Though they do not explicitly suggest that university researchers ofteaching ought to pack their bags, their enthusiasm for teacher researchers,combined with their frustration with the domination of teaching research byinstitutions of higher education, certainly suggests that university-basedresearchers should rethink the nature of their work. The question of whoproduces knowledge of teaching evokes memories of some of the criticism I, as auniversity-level scholar, received on venturing into this arena. A momentaryautobiographical reference offers what I trust will be some further illumination onthe important distinction between who makes and who consumes teacherknowledge.

In preparing an article for the third edition of the Handbook forResearch on Teaching(Fenstermacher, 1986), I tried to work out therelationship between social science research on teaching and the uses thatteachers might make of this research. In doing so, I found it helpful to draw adistinction between knowledge production and knowledge use, contendingthereafter that scientists produced knowledge and teachers used it. For thisconstruction, I was roundly, and I now believe rightly, criticized by some of myCanadian colleagues who offered Schn's work as a counter to mine (seeFenstermacher, 1987; Kilbourn, 1987; Munby, 1987; and Russell, 1987).Teachers, they retorted, are more than simply passive consumers of knowledge;they are producers of it as well, and users of their own knowledge as well as that

produced by others. Cochran-Smith and Lytle make similar contentions, buildingup the theoretical rationale for these contentions as well as exploring methodsand procedures for teachers becoming engaged in their own research.

At the time I made the producer/user distinction, I was thinking of teachersas engaged in a quite different form of inquiry from that employed by academicresearchers. That inquiry, I argued, was practical, in contrast to the formalinquiry of university researchers. I will, with but a few exceptions, continue todefend this position. However, what I did not contend then, but would do now,is that practical inquiry yields a form of knowing just as formal inquiry does.Additionally, I believe the forms of inquiry employed by university-level

researchers and elementary and secondary school teachers are typically different,though they need not be differentiated by occupational role (Richardson, inpress/a, provides an extended discussion of formal and practical inquiry,indicating how both may be done by either researcher or practitioner). That is, ateacher may engage in formal or practical inquiry, depending on the problemstudied, the methods employed, and the intended scope of the results. Aresearcher may as well engage in either form of inquiry.


18/61


19/61

19

With these remarks in answer to the fourth question, the overview ofresearch programs may be brought to a close. Four questions were posed andanswered, in ways that led to the identification of two major types of knowledge,

formal and practical. Not all research programs discussed fall neatly into thesetwo knowledge categories, as not all of them are efforts to specifically addressepistemological issues. Several of the programs that are not especially sensitiveto or concerned with epistemological issues engender ambiguity about whattypes of knowledge they are striving to study or produce. This ambiguity may beunproblematic, though we will soon see that it may also be quite the reverse.One cannot make much more progress in understanding the epistemologicalissues, or in examining the conceptions of science upon which these researchprograms are based, without considering epistemology itself.

III. CONSIDERATIONS FROM EPISTEMOLOGYThe preceding overview identified two primary conceptions of knowledge,

formal and practical. In this part, these two conceptions will be examined indetail by connecting them to considerations from that branch of philosophy calledepistemology. Many people regard this area of philosophy as preoccupied withquestions of what we know and how we know it. True enough, though it alsoincludes such topics as causation, empiricism, intuition, reasoning, sense data,and universals. One of the most succinct definitions of the field is offered by D.W. Hamlyn (1967, pp. 8-9): "Epistemology, or the theory of knowledge, is thatbranch of philosophy which is concerned with the nature and scope of

knowledge, its presuppositions and basis, and the general reliability of the claimsto knowledge."

Epistemologists differ on how many different types of knowledge theremight be, though there is a clear fondness--as in most academics--for the viewthat there are three types. David Pears (1971) makes the common tripartitedistinction among, (1) factual or propositional knowledge, (2) knowledge of howto do things, and (3) knowledge by acquaintance. Keith Lehrer (1990a) describesa similar division as he examines different senses of the expression "to know,"ranging from (1) recognizing something as information, to (2) having some specialform of competence, to (3) being acquainted with someone or something. Albert

Jonsen and Stephen Toulmin (1988, pp. 58-59) credit Plato and Aristotle for threemain types of knowledge: "theoretical comprehension of abstract arguments in thesciences, practical command of general craft techniques, and the prudentialwisdom that is required to handle particular legal problems or medical cases." Inaddition to these more or less standard categories, there is the tacit knowledgedescribed by Michael Polanyi (1964, 1966), which is often considered a sub-species of one or more of the three categories mentioned just above.


20/61


21/61

21

epist m and techn . Since Aristotle, the term


22/61

22

"standard analysis" of knowledge (Shope, 1983). It has been succinctly describedby the philosopher, Edmund Gettier (1963), who is frequently credited withdealing it a near-knock-out punch. The standard analysis reads as follows:

Sknows that p, if and only if (i) pis true

(ii) Sbelieves that p, and (iii) Sis justified in believing that p.

(Gettier (1963, p. 121)

The standard analysis establishes strict conditions for a knowledge claim.The claim must be true, it must be believed by the claimant, and the claimantmust be justified in believing the claim. Thus where prefers to the proposition,"The earth has a spherical shape," in order for some person Sto be said to know

p, it must be the case that, (i) it is true that the earth has a spherical shape (thiscondition is established within the discipline that concerns itself with propositionsof this type), (ii) Sbelieves that the earth has a spherical shape, and (iii) Shas

justification for believing that the earth has a spherical shape.

The standard analysis has sometimes been called the "justified true belief"or JTB account of knowledge. Gettier's (1963) critique posed a series ofcounterexamples to the standard analysis that led to number of attempts to repairthe weaknesses he exposed. Much of the ensuing concern has been with themeaning of truein the statement "xis true," and with what constitutes a proper

justification for S's believing that p(see, for example, Chisholm, 1989, Lehrer,

1990a & b, Pappas & Swain, 1978, Shope, 1983). It would be a seriousmisunderstanding of the consequences of Gettier's counterexamples to infer thatthere is no longer a satisfactory or commonly accepted conception of what itmeans to know something in the propositional or informational sense of the term.A large number of contemporary epistemologists are in agreement that we canhave knowledge of the world, that there are specific conditions that must be metif our claims to knowledge are to succeed, and that some form of justification isrequired as one among these conditions. The issue of justification, inepistemology, is over the particulars of what constitutes a defensible justification,not over whether justification is required. It will always be the case, as near ascan be foretold, that some form of justification will be required if we are to know,

as Lehrer (1990b, p. 253) puts it, "that the information one receives is correct."

Gettier's critique and a host of related analyses have put an end toconceptions of formal knowledge as pristine, permanent, and absolute. Again,however, it would be an error to conclude that because our knowledge is fallible,it is therefore useless to worry about what it means, or that it is simply tooconfusing to take seriously any epistemological account of knowledge. Fallibilism,the view that we can be wrong about we claim to know, is a fairly well accepted


23/61

23

doctrine among contemporary epistemologists; it is not regarded as devastating inits consequences for a well-wrought theory of knowledge. Nor can we ignore thedemands of justification on the grounds that do not have the rock-solid, gold-

plated conceptions of truth that once graced theories in epistemology.

Consider Anthony Quinton's (1967, p. 346) reaction to the claim that ifknowledge entails truth, we can never know anything:

This objection is misconceived. If I firmly believe that something istrue on what I take to be sufficient grounds, I am right to say that Iknow it. It may be that the grounds are, in fact, insufficient and thatwhat I claim to know is false. In that case my claim is mistaken, butit does not follow that I was wrong to make it in the sense that Ihad no justification for doing so.

Although he occupies a position in epistemology that departs in severalrespects from the position I have taken in this chapter, Roderick Chisholmprovides a helpful way to think about issues of justification and truth. He setsforth what he calls "The 13 Steps" for considering the epistemic status ofknowledge claims (Chisholm,1989, p. 16), as follows:

+6. Certain+5. Obvious+4. Evident

+3. Beyond Reasonable Doubt

+2. Epistemically in the clear+1. Probable

0. Counterbalanced-1. Probably False

-2. In the Clear to Disbelieve-3. Reasonable to Disbelieve

-4. Evidently False-5. Obviously False-6. Certainly False

These steps provide a good feel for how epistemologists look at what I am

calling formal knowledge. Although for Chisholm a proposition cannot beknownuntil it attains the fourth, positive step (Evident) or above, otherepistemologists would entertain a proposition's having epistemic merit at the levelof "Beyond Reasonable Doubt." At the fourth, or Evident, step, the propositionmeets modified standards for what is known as justified true belief. For otherepistemologists, a claim is sufficiently justified to count as knowledge if it isreasonably believed by the holder, the holder having sufficient evidence toestablish the claim against its competitors. This notion is parallel to the idea of


24/61

24

"objectively reasonable belief" described by Green (1971) in his analysis ofknowledge and belief in education.

Some epistemologists doubt whether objectively reasonable belief countsas knowledge. Oliver Johnson (1980) argues for a distinction between reasonablebelief and knowledge, concluding that we should "reserve the term 'knowledge' forbeliefs whose truth we can justify" (p. 126), and use the expression "reasonablebelief" in cases where we cannot justify the truth of the proposition but "for whichwe can offer reasons that we believe in some sense to be 'good'" (p. 126). Thisview is consistent with Chisholm's, in that what is just below the level of Evidentis not knowledge, though it is based on sufficient reasons as to be beyondreasonable doubt. However, there is some debate on the matter of when (andhow) a proposition flips from being "reasonable belief" (or, as Green would say,"objectively reasonable belief") to being properly identified as knowledge. Without

becoming mired in this debate, it is sufficient for our purposes (and in accord witha fair body of epistemological thought) to hold that a proposition is known by itsholder if the holder believes the proposition and has evidence to establish itsreasonableness in relation to other, competing claims.

Objectively reasonable belief may not be a sufficiently strong standard tosustain the concept of knowledge as it is currently used in the social or naturalsciences (though this point is also subject to debate), yet if this standard is met inteacher knowledge claims there are few who would quibble over half a notch orso on Chisholm's scale. An even "softer" version of standards for knowledgestipulates that what counts as knowledge in a given context is relative to what is

already known in that context. If an assertion is about matters of which little isknown, we may be entitled to say that we know p, with only modest justification.In other words, "It's the best we've got" at this time and so we are entitled to claimto know it. In areas where more is known, our obligation to address this evidenceand consider it in relation to our own grounds for believing the proposition iscorrespondingly increased. No softening of the standards may, however, golower than Chisholm's "epistemically in the clear," by which is meant that theholder can show that there are not better grounds for rejecting the propositionthan for accepting it. Clearly this is a weak standard; any claim to know xon thisstandard is highly suspect, and would have to be made in a context where therewas little or no evidence available to guide consideration of the proposition.

In summary, the analysis of propositional knowledge offers some latitudefor interpretation, but does not permit definitional anarchy. It its strict form,knowledge continues to be understood as a form of justified true belief, thoughwithout a sense of the absolute, permanent qualities that once attached to thisnotion. A more modest standard permits us to make knowledge claims on thebasis of objectively reasonable belief. An even milder form permits us to gaugeour knowledge claims in relation to the evidential character of the context in


25/61

25

which the claim is made, thereby "allowing" one to claim to know somethingwithout meeting the standard criteria for objectively reasonable criteria. In myown case, I believe that objectively reasonable belief is an acceptable form of

knowledge within the context of educational practice (though it may not satisfythe canons for educational research, at least not in the more conventional scienceconceptions of educational research). The even milder version, what might becalled the contexted knowledge version, is feasible, but should be employed withgreat caution.

Given the view of propositional knowledge presented here, one can seehow the following definition of knowledge, taken from a recent article in theReview of Educational Research, might bruise the sensitivities of anepistemologist:

For researchers in the field of cognition and literature, it goes nearlywithout saying that knowledge refers to an individual's personalstock of information, skills, experiences, beliefs, and memories. Thisknowledge is always idiosyncratic, reflecting the vagaries of aperson's own history.... In the literature we are reviewing here,knowledgeencompasses all that a person knows or believes tobe true, whether or not it is verified as true in some sort of objectiveor external way. (Alexander, Schallert & Hare, 1991, p. 317;emphasis in original).

The authors argue that "this use of the term knowledgecontrasts with the use

of the term in the field of epistemology, where knowledgeoften refers tojustified true beliefs and is reserved for universal, or absolute, truths" (p. 317). AsI hope is now clear, the concept of knowledge within contemporary epistemologyhas a far broader range than is presumed by these writers. However, the positionthey identify does raise a noteworthy distinction between the way the termknowledgemight be used in, say, social science and its use in epistemology.More will be made of this distinction in a later section.

B. The Character of Performance KnowledgeWhat epistemologists have generally called "knowing how," skill

knowledge, or competent performance, is here referred to as performance

knowledge. It is typically considered to be quite distinct from propositionalknowledge. Many of us, on hearing the notion of performance knowledge, thinkof the difference made famous by Gilbert Ryle (1949) between knowing that andknowing how. Of course the distinction predates Ryle by a few thousand years,though Ryle is generally credited with reintroducing the topic into Anglo-Americanphilosophy. In its Greek form, it was, as already noted, discussed as techn .


26/61

26

As such it was knowing how to make something, a skill or capacity for theexercise of some craft. Building a house, playing the lute, or translatinglanguages are good examples.

Mark Johnson, a philosopher on whose work Connelly and Clandinin basetheir notion of embodied knowledge, argues for a view of knowledge that "growsout of one's personal experience and is the very means of transformation of thatexperience" (M. Johnson, 1989, p. 364). Johnson then argues that seen in thisway, "personal practical knowledge is a quite similar to the classical Greekconception of techn " (p. 364). Elaborating on the connection between technand personal practical knowledge, Johnson writes:

The association of personal, practical knowledge withtechn broadly defined is notmeant to suggest that we are

dealing only with a form of "know-how." On the contrary, followingDewey, I wish to reject the whole dichotomy between "knowingthat" ... and "knowing how" .... No good has come from this allegedcrucial epistemological distinction, which has served only to give usa picture of ourselves as cognitively fragmented, and which hasreinforced the ill-conceived "theory/practice" dichotomy. (M.Johnson, 1989, p. 365)

My own sense is that only a philosopher, and then only a few of them,could feel "cognitively fragmented" by the distinction between knowing how andknowing that. Clearly our use of everyday language acknowledges a difference

between the two, as most people deal differently with a claim like "I know thatsmoking is the leading cause of lung cancer" than they do with a claim like "I knowhow to smoke a pack of cigarettes all at once." To the former claim we say, "Whatevidence do you have to substantiate such a claim?" while to the latter, we say,"Oh, really! Show me." Indeed while Johnson appears to opt for collapsing thetwo forms into some new category, much of the philosophical discussion of theknowing- how/knowing-that distinction has been on whether knowing how isdependent on knowing that for its for its justification.

Because a considerable portion of the TK/P literature is concerned with the"know how" that teachers accumulate through experience and reflection, it is

worth pursuing the distinction a bit further. Ryle (1949) provides the best known,if not the most explicit modern discussion (though Israel Scheffler, 1965,provides an extensive analysis of this and related concepts as they bear on theconcept of education). Concerned with what he regarded as fundamental errorsin the Cartesian view of the relationship of mind and body, Ryle set out to offer acorrection to the dualist thesis (that mind and body are separate, but interrelated).This correction involved, among other things, arguing that knowing how couldnot be reduced to knowing that. In one of the conclusions he draws on this point,


27/61

27

Ryle (1949, p. 30) states, "It is therefore possible for people intelligently to performsome sorts of operations when they are not yet able to consider any propositionsenjoining how they should be performed." If it were in fact necessary, argues

Ryle, to consider such propositions, it would create a most confounding difficulty:"If for any operation to be intelligently executed, a prior theoretical operation hadfirst to be performed and performed intelligently, it would be a logical impossibilityfor anyone ever to break into the circle" (p. 30).

Ryle's compelling prose and evocative arguments have convinced manyreaders that knowing-how and knowing-that are distinct domains, independentof one another. On the other hand, there is some consideration in thephilosophical literature that knowing how is, in an important sense, dependent onknowing that. For example, in A. J. Ayer's (1956) well-known work, TheProblem of Knowledge, he states that under certain conditions "we can

construe knowing how to do things as being, in its fashion, a matter of knowingfacts" (pp. 13-14). More recently, Lehrer (1990a, p. 4) remarked that "it is oftenaffirmed that to know something in the other senses of


28/61

28

performance knowledge is every bit as important to its epistemic status as it is inthe case of propositional knowledge, and that such justification is not simply inthe performance of the skill or the craft but also in establishing the

reasonableness of the performance and the evidence connecting its purpose to itseventual outcome. With this stipulation in mind, it is time to address theconnections between propositional and performance knowledge, on the one hand,and formal and practical knowledge, on the other.

C. Distinguishing propositional knowledge from TK/F and performance knowledge from TK/P

It is tempting to contend that what epistemologists call propositionalknowledge is the same as what I have identified as Teacher Knowledge: Formal,and that performance knowledge is the same as Teacher Knowledge: Practical.While there are clear affinities between the categories, it is an unwise move to

conflate formal with propositional, and practical with performance. One obviousreason is the lack of precision in the concept of practical knowledge, as it is usedacross the several teaching research programs. Even though a significant shareof this concept might be explained by the epistemological notion of performanceknowledge, far more falls with the domain of TK/P than is accounted for in thephilosopher's concept of performance knowledge.

Furthermore, as the notion of TK/F is developed here, there are strongconnections between certain methods of inquiry (i.e., scientific research) andclaims to formal knowledge. Thus as developed in this chapter, there are certainstandards of significance, generalizability or validity (in the broadest sense of

these terms) that must be met to properly claim that one is in possession forformal knowledge. These standards are introduced here in part because thischapter is about teaching research. Thus to properly claim that one is inpossession of formal knowledge, one's knowledge claims must be justified in sucha manner that they range beyond the immediate context, situation, or slice oftime. In the epistemological literature, a straight propositional claim might behighly constrained by time and context, and still meet epistemological criteria forcounting as knowledge. That is, a proposition may be quite specific or timebound, yet still pass epistemological muster as knowledge. In contrast, what Ihave called formal knowledge will not survive epistemological scrutiny asknowledge if it is restricted in its application to time, place, and situation.

Practical knowledge, on the other hand, is bounded by time, place orsituation. To claim to know something practically is to claim to know somethingabout an action, event or situation in this particular instance. Practicalknowledge consists in some part of what epistemologists call performanceknowledge, though it ranges well beyond this limited notion of performance orcompetence (to include, for example, what Jonsen and Toulmin (1988) callprudential wisdom, or what Shulman (1986b) calls strategic knowledge). Thus


29/61

29

even though the notion of TK/P is in large measure rooted in the epistemologicalconcept of performance knowledge, it is often discussed in the literature as amuch broader, more inclusive concept.

With these connections in mind, the critical epistemological point can bemade: Both TK/F and TK/P are subject to evidentiary scrutiny if there are to countas knowledge in any useful sense of the term. That is, to claim that we knowsomething about teachers or teaching, or that teachers know something aboutteaching, it must be the case that we can be in error, that we understand whatwould count as showing that we are in error, and that we have a basis forestablishing that we have not made an error. That we claim to have practicalknowledge does not relieve us of the obligation to show how it is objectivelyreasonable to believe what we are contending. Justification or warrant are asvital to TK/P claims as they are to TK/F claims, though they are different

undertakings depending on the domain in which one stakes one's claim. Whatdistinguishes this effort in the practical, as opposed to the formal domain, is thatwe do not require the methods or other paraphernalia of science to do this.

Part IV of this review, covering the scientific study of teaching, providesfurther discussion of the differences between formal and practical knowledge inresearch on teaching. Deferring this discussion until later in this chapter permitsus to move on to another critical epistemological feature, the relationship betweenknowledge and belief.

D. Knowledge and beliefIt is an eye-opening experience for a student of epistemology to note how

knowledge and belief are grouped together in so many accounts of teacherknowledge. As I hope is now evident, a claim to know something isepistemologically different from simply having a belief in something. Consider, asan example, the proposition, "Harry's car is parked in garage A, lower level." If youwere to probe the epistemic status of this claim, you might get one of these threedifferent responses:

(1) I saw his car parked there this morning when I parked my car, and his secretary says she still has the car keys he left with her when he arrived.

(2) Harry always parks his car on the lower level of A.(3) Guys like Harry prefer the lower level of A.

Response (1) has a higher epistemic status because it is characterized bythe evidence of actually seeing Harry's car on level A and having the testimony ofa reliable person that Harry could not have moved his car since it was observed inthe garage. Claims (2) and (3) exhibit diminishing epistemic status due to the fall-off in evidentiary support and the increasing difficulty of setting out a convincing


30/61

30

justification. However, in all three instances there is no prima faciedifficulty insaying that the person believesthat Harry's car is parked in the lower level of A,whereas there is an increasingly serious problem, as one moves from (1) to (2) to

(3), with saying that the person knowsthat Harry's car is parked on the lowerlevel of A.

This distinction, so commonly drawn in epistemology, is not so commonlyused in research on teacher knowledge. Sometimes this state of affairs isunproblematic, as when the term `knowledge' is being used as an inclusive groupname. This appears to be the case for the use employed by Patricia Alexander,Diane Schallert, and Victoria Hare (1991), quoted a few pages earlier, whenstating that knowledge"refers to an individual's personal stock of information,skills, experiences, beliefs, and memories" (p. 317). In this grouping orcategorizing sense of the term, to call all these cognitive events or mental states

"knowledge" is not to make any claims about their epistemic status. The groupingor categorizing sense of the term is consistent with such dictionary definitions ofknowledgeas "familiarity, awareness or understanding gained throughexperience or study; the sum or range of what has been perceived, discovered, orlearned" (The American Heritage Dictionary, 3rd ed). This groupingsense may also be the basis for Dona Kagan's (1990) clustering of knowledge andbelief, although this case is less clear than the Alexander, Shallert and Hareexample. Kagan states that she often uses "beliefand knowledgeinterchangeably .... I do so in light of mounting evidence that much of what ateacher knows of his or her craft appears to be defined in highly subjective terms..." (1990, p. 421).

This is a puzzling contention, for it seems to acknowledge that knowledgeand belief might be distinguished by notions of objectivity and subjectivity,respectively, yet the decision here appears to be treat them as having the sameepistemic quality becauseof the apparent subjectivity of what teachers "know."I believe that Kagan would be on firmer ground were she to contend that theterms knowledgeand beliefare sometimes used in some research programsas if they were identical in meaning and epistemic status, and thus she will usethem interchangeably. However, the moment she bases the decision to use theterms interchangeably on the notion of subjectivity, she raises the general issue ofepistemic import. It is precisely in this event that the distinction between

knowledge and belief must be drawn.

Given what appears to be a growing tendency in the educational researchliterature to discuss knowledge and belief as if the terms were synonymous, thereis much to be gained from making a distinction between using knowledgeas agrouping or classifying term, on the one hand, and using it to imply or entail therelative epistemological merit of one's claims, on the other. When used as agrouping term, we may with impunity, include all sorts of phenomena under the


31/61


32/61

32

E. Local Craft Situated and Tacit KnowledgeThere are a number of permutations on the formal/practical distinction that

frequently appear in the teacher knowledge literature. Understanding where and

how these fit into epistemological discourse is essential if one is to have a graspof how these variant forms of knowledge bear on the analysis of this chapter.Because my purpose is only to suggest how these variant forms fit into theframework of this chapter, my accounts will be too brief to acknowledge both theobvious merits and possible shortcomings of this work.

Though the concept of local knowledge is discussed by Geertz (1983) anda number of other anthropologists, the application of this concept to teaching iswell illustrated in a multi-year study of reading achievement among Spanish-speaking children by Goldenberg and Gallimore (1991). In this work a distinctionis drawn between local knowledge and research knowledge. Local knowledge is

defined in a way that appears to place it within the category of TK/P. ForGoldenberg and Gallimore, local knowledge is direct experience with how thingswork. It is, they argue, prior to research knowledge, which "cannot have a directbearing on practice because it is oblivious of compelling local issues that framethe thinking and drive the behavior of practitioners in a particular locale" (p. 2).They go on to argue that the success of school reform depends on a betterunderstanding of "the interplay between research knowledge and local knowledge"(p. 2). The Goldenberg and Gallimore research indicates that research knowledgeplaces no great demands on practitioners or policy makers, "no matter how validor compelling such knowledge is to researchers" (p. 11). It was the localknowledge of the practitioners that determined whether new approaches to

literacy and language acquisition, as established by research, would be adoptedby the schools.

The contrasting of local and research knowledge is a good example of thedistinction I have drawn between TK/P and TK/F, wherein local knowledge is avariety of the practical, while research knowledge is a variety of the formal. Thelocal/research knowledge distinction also makes clear why I have avoidedconnecting formal with propositional, and practical with performance. In thisstudy, the local knowledge was clearly propositional, though it may also haveincluded performance knowledge. Though propositional in form, it was notresearch knowledge. Local knowledge can be propositional, explicitly stated,

tested, and fully justified in a manner similar to any scientific claim; the soledifference is that its range of application is restricted to the context in which itwas formed. In the Goldenberg and Gallimore context, local knowledge appearsto be knowledge of particular instances, in contrast to general knowledge, whichis knowledge of phenomena that share similar properties.


33/61

33

Craft knowledge, like local knowledge, appears to reside in the TK/Pcategory. Though teaching as a craft was explored in considerable depth by Tom

(1984), he was more interested in developing an overall conception of theoccupation of teaching than in teasing out the epistemic features of the concept ofcraft knowledge. That effort would await more recent treatments, such as thoseby Gaea Leinhardt (1988, 1990). Leinhardt's work, like other studies of practicalknowledge, is grounded in a high degree of regard for teachers. HoweverLeinhardt is careful not to regard all that they say or do as worthy ofacknowledgement as wisdom or knowledge. From these twin perspectives ofhigh regard and due caution, she formulates the notions of craft and situatedknowledge.

Situated knowledge is "a form of expertise in which declarative knowledge

is highly proceduralized and automatic and in which a highly efficient collection ofheuristics exist for the solution of very specific problems in teaching" (Leinhardt,1988, p. 146). Leinhardt contends that situated knowledge can be contrastedwith "context-free principled knowledge that is applicable or accessible in anycircumstance.... Principled context-free knowledge is detached and generallytrue. It will always work and is thus very powerful..." (1988, p. 148). Situatedknowledge competes with principled knowledge, such that when the two areside-by-side, "situated knowledge will be preferred in the particular situation inwhich it normally operates" (Leinhardt, 1988, p. 148). Note the similarity betweenthis characteristic of situated knowledge and Goldenberg and Gallimore'scontention that, unless specific steps are taken to address the possibility, local

knowledge will almost always take precedence over research knowledge.

A few years after working out her concept of situated knowledge, Leinhardtturned to the problem of assessing teacher knowledge, arguing that anyworthwhile assessment would have to take the craft knowledge of teachers intoaccount. Though she did not, in this more current version, compare her earlierdefinition of situated knowledge with the later notion of craft knowledge, onesenses a high degree of overlap. Craft knowledge includes "deep, sensitive,location-specific knowledge of teaching," as well as "fragmentary, superstitiousand often inaccurate opinions" (Leinhardt, 1990, p. 18). In her view, thesuccessful assessment of teacher knowledge and expertise calls for merging craft

knowledge with theory and empirical research. Contentions such this one pointto a fairly sharp distinction between craft or situated knowledge, on the one hand,and research knowledge on the other.

Another formulation of craft knowledge is offered by Grimmett andMacKinnon (1992). Their view departs from more descriptive characterizations,


34/61

34

such as Leinhardt's, constituting instead a more prescriptive sense of craft inteaching. This normative sense of craft is clearly evident in the followingformulation of the concept:

Craft knowledge of teaching is not substantive, subject matterknowledge .... It is a particular form of morally appropriateintelligent and sensible know-how that is constructed by teachers,holding progressive and radical educational beliefs, in the context oftheir lived experiences and work around issues of content-relatedand learner-focused pedagogy. (Grimmett & MacKinnon, 1992, p.396)

The authors identify teachers who possess such knowledge as crafty teachers, "inthe dexterous, ingenious sense" of the term (p. 429).

There is obviously some variation in the literature on what we are tounderstand by the notions of craft and situated knowledge, ranging fromLeinhardt's attempts to describe these forms of knowledge as they occur inteaching of various subjects to Grimmett and MacKinnon'sconception of what we ought to understand when referencing the concept of craftknowledge. Another, more philosophical view of these concepts is offered byRobert Orton (in press).

Orton distinguishes between teacher knowledge that is situated and tacit.Situated knowledge is restricted in its range of application by place, time or

context, much as Leinhardt contends. However, for Orton the fact that it issituated does not mean that it is incapable of discursive expression (theepistemologist would say that it can be articulated). Tacit knowledge, in contrast,cannot be articulated; it is "read off" or inferred from the performance of some skillor ability. Orton argues that situated knowledge does not present majorchallenges to epistemological justification, but tacit knowledge does. Because itcan only be inferred from performance, tacit knowledge resists the kind of

justification that would permit us to properly identify it as knowledge. Orton'sanalysis offers a number of provocative perspectives on the notions of situatedand tacit knowledge. We will have occasion to reconsider some of his ideas atthe end of part IV. Before moving on to that part, we will look one more time at

the topic of justification.

F. The justification of knowledge claimsIn their analysis of the cultures of teaching, Feiman-Nemser and Floden

(1986, p. 515) write that it does not follow "that everything a teacher believes or iswilling to act on merits the label knowledge, although that view has somesupport. Such a position makes the concept of knowledge as justified beliefmeaningless." Their point is well-taken, though there are reasons why the term


35/61

35

knowis preferred over related terms. Many philosophers regard the termknowledgea "purr" word; it feels so good to use and hear that we almost purrwhen doing so. It is also a concept with legitimating qualities; everyone has

beliefs and opinions, but knowledge is something special, something that elevatesone's thoughts and expressions beyond "mere" belief or opinion. Consider thesecomments by Elbaz on why she undertook her study of Sarah:

My intention was not simply to present for its own sake a picture ofthe teacher as a holder of knowledge. Rather I assumed that tochange our view of the teacher is to open up new possibilities foraction. Further, since knowledge is power, to see oneself as holdingknowledge is also to see oneself differently in relation to existingsources of power. (Elbaz, 1983, p. 165)

One finds similar reflections in Clandinin, Shulman, Cochran-Smith and Lytle, andother researchers on teacher knowledge. While the sentiment behind such usesof the term knowledgemay be most noble, they are ultimately destructive oftheir purpose when unaccompanied by ways of distinguishing between claimsthat have little or no epistemic status and those that have a great deal (MargaretBuchmann has made this point on a number of occasions; see, for example,Buchmann, 1984, 1986, 1987, 1993).

There are, as a result of this discussion, a number of possibilities for usingthe term `knowledge' when referring to the mental states and activities ofteachers::

1. It is indeed a type of knowledge, either formal or practical, and hasconsiderable epistemic import.

2. It is simply a generic name to describe a broad range of mentalstates of teachers that arise from their training, experience andreflection, and has little if any epistemological import.

3. It is deliberate selection of a socially valued term, used with theprimary intention of legitimating the insights,understandings, beliefs, and so forth that teachers possess.

Note that the third possibility does not mention anything about

epistemological import. That is because epistemic import is not a consideration inthis third conception. However, the only way version 3 succeeds is if it issomehow connected to version 1; i.e., that the notion of "teacher knowledge" hasepistemological import and is not simply a group name designating thecognitions, mental states, or understandings of teachers. Though there is nothingintrinsically wrong with version 2 (it does, as noted earlier, follow one or moredictionary definitions of knowledge), it is unconnected to epistemic merit orstatus. That is, there is no basis in version 2 for deciding whether the knowledge


36/61

36

of one teacher or researcher is better than, more trustworthy than, less troubledby error, more resistant to objection and criticism than the knowledge of any otherteacher or researcher. To the extent a conception of knowledge does have

epistemic merit, it will provide a basis for determining the strength, confidence, ortrustworthiness of a claim to know something, whether that knowledge ispropositional or performative, formal or practical.

To put these last few points another way, the claim that teachers knowsomething is made either with or without epistemological import. If madewithout such import, it is uninteresting as an epistemological claim for its makesno assertions about knowledge, per se. Rather it is simply a taxonomic aid to theresearcher, who seeks to group, as the dictionary states, "familiarity, awarenessor understanding gained through experience or study." If such claims are madewith the intent of having epistemic import, then there must be a way to justify the

claims. Such justificatio

Documents

Fenstermarcher Gary_The Knower and the Known